Worldwide Historical Child Stunting Dataset

This dataset presents child stunting rates from the systematic review of historical child growth studies performed by Schneider et al. (2026). The Decline of Child Stunting in 122 Countries: A Systematic Review of Child Growth Studies Since the Nineteenth Century. BMJ Global Health, forthcoming. Full details for how the data were collected and how stunting rates were computed is available in the supplementary materials of the article, which are also included in this deposit. Please note the following important issues with the data before reuse: Each case in the data is a stunting estimate for a particular study. Some studies produce multiple cases because they provided estimates for multiple years. The underlying studies are heterogeneous and most are not nationally representative. Each study has a series of certainty of evidence scores that code issues related to study quality. These are discussed at length in Appendix I. The extent to which trends for any particular country should be trusted is dependent on the quality of the underlying data. We therefore do not recommend using the data for simple panel regressions that would not account for the underlying heterogeneity of the data. The stunting rates in the dataset are for children age 2 to 10 rather than the traditional ages 0 to 5. Thus, in order to merge this data with the Joint Malnutrition Estimates (JME) Dataset from Unicef, the WHO and the World Bank, future researchers will need to use the stunting rates for ages 2 to 5 reported in their data. Instructions for and discussion of merging this data with the JME data is presented in Appendix H. Bibliographic information is provided for all studies, but many are not available digitally and will need to be consulted in libraries. Please direct all queries to Eric Schneider, London School of Economics and Political Science, e.b.schneider@lse.ac.uk, ericbschneider.com.